LoginStart beta
Ayneye
How it worksStart beta

World-State

World-State is the product interface between video and software.

Raw video is expensive to search repeatedly. World-State turns footage into structured objects, events, zones, timestamps, relationships, confidence, and evidence references that software can inspect.

World-State data model

PartPurposeExample fields
ObjectsRepresent things observed over timeobject_id, label, confidence, first_seen, last_seen
EventsRepresent changes or actionsevent_type, timestamp, participants, evidence_refs
ZonesRepresent spatial contextzone_id, name, geometry, camera_ref
RelationshipsConnect objects and eventssubject, relation, object, confidence
Review flagsExpose uncertaintylow_confidence, ambiguous_scene, human_review_required

Why this matters to product teams

World-State lets teams build dashboards, search, APIs, alerts, and agents over a stable representation. It does not make every visual problem solved. It gives the product a durable memory layer that can be tested, versioned, cached, and compared against evidence. That is what makes repeated queries cheaper and debugging easier.

World-State quality checklist

  1. Important objects appear with stable IDs.
  2. Events include timestamps and evidence_refs.
  3. Low-confidence observations are not hidden.
  4. Ambiguous events become review_required.
  5. Updates are versioned so agents know what state they read.

Product boundaries are part of the product, not footnotes. Ayneye is not presented as a replacement for every GPU video foundation model, not a free-infinite-query engine, and not an autonomous surveillance decision system. The current beta path is controlled signup, dashboard, REST API, bounded processing, evidence artifacts, visible limits, and read-only agent/MCP-style evaluation. Hard scenes, identity-sensitive workflows, emergency response, physical access, discipline, and destructive actions require human review or remain blocked.